hello
we host 8 rds servers (windows server 2008 r2) on 4 hosts (esxi 5.0.0. 2312428), 1 vcenter server (5.0.0. 623373). all this is hosted in claranet's data centre on what they call vplatform.
our issue as of late is - we are seeing users remote session slow down, get frozen across multiple sites. checks on rds servers indicate no windows issues nor problems with available cpu or memory. call to claranet to investgiate potential IOPs issue but they couldnt find anything on the SAN (which is shared amongst other customers of claranet). they did however report that we had 'quite a lot of bandwidth spikes on datatstore 1. we have 4 datastores but my colleague has setup the C: drives of all 25 virtual machines on datastore1. i cant find anything that sticks out on the performance charts.
im not sure if setting up a datastore like this is a bad idea, especiall with 25vm c: on the same datastore. i've attached a graph of the bandwidth spikes sent by claranet.